分析员工画像
gethouston2026-10-24查看 495复制 795
将我了解的关于某位员工的所有信息汇总成一页纸——HR 档案、入职计划、近期一对一沟通、面试流程历史。在一对一沟通、薪酬谈话或艰难会议之前很有用。
📘如何使用这个技能
- 1复制技能内容·点击上方「Copy SKILL.md」复制全部内容,或 Download 下载到本地
- 2打开你的 AI 客户端·支持 Claude Code、Trae、Cursor、ChatGPT、DeepSeek、Kimi 等任意 AI 助手
- 3粘贴为系统提示/知识文件·Claude Code:放入 .claude/skills/ 目录;普通对话:作为首条消息发送并说「请严格按这份技能文档执行」
- 4直接提出你的需求·AI 会按技能定义的工作流、格式和约束来处理,无需再解释背景
💡 提示:技能生效后,每次对话只需一句需求,AI 会按 SKILL.md 的流程执行。
SKILL.md5169 chars
---
name: profile-an-employee
description: "Pull together everything I know about an employee into a one-pager - HR profile, onboarding plan, recent check-ins, interview loop history. Useful before a 1:1, a comp conversation, or a hard meeting."
version: 1
category: People
featured: no
image: busts-in-silhouette
integrations: [notion, slack, loops]
x_houston:
created_by: houston
skill_schema: 1
---
# Profile an Employee
## When to use
- Explicit: "tell me about {employee}", "pull everything on {employee}", "prep me for my 1:1 with {employee}", "dossier on {employee}".
- Implicit: routed before review cycle, sensitive conversation (performance improvement plan, promotion, comp change), or exit interview.
- Frequency: on-demand per employee.
## Connections I need
I run external work through Composio. Before this skill runs I check that the categories below are linked. Missing → I name the category, ask you to connect it from the Integrations tab, stop.
- **HR platform (Gusto, Deel, Rippling, Justworks)** — read role, level, tenure, manager, comp, work authorization. Required.
- **Chat (Slack)** — pull recent thread context if relevant. Optional.
- **Docs (Notion)** — pull review notes or 1:1 docs if you keep them there. Optional.
- **Inbox (Loops)** — pull recent comms if useful. Optional.
If your HR platform isn't connected I stop and ask you to link Gusto, Deel, Rippling, or Justworks from the Integrations tab.
## Information I need
I read your people context first. For every required field that's missing I ask ONE plain-language question (best modality: connected app > file drop > URL > paste) and wait.
- **Employee identity** — Required. Why I need it: I won't pull a dossier for someone I can't pin down. If missing I ask: "Which employee, full name, and team if you have it?"
- **Authorization to view confidential fields** — Required when comp or visa data is requested. Why I need it: I never leak one employee's comp or status to another without it. If missing I ask: "Should this dossier include comp and work-authorization details, or keep it to role and tenure only?"
- **Roster source** — Required when HR platform isn't connected. Why I need it: I need somewhere to read the basics. If missing I ask: "Connect your HR platform so I can pull this directly, or paste the employee record."
## Steps
<!-- houston-workflow:v1 -->
1. **Read people-context doc.** Read `context/people-context.md` for leveling, comp bands, confidentiality rules around dossier content. If missing/empty, tell user: "I need the people-context doc first - run the set-up-my-people-info skill." Stop.
2. **Read ledger.** `config/context-ledger.json`. If HR platform not connected and no roster link recorded, ask ONE targeted question with modality hint - "Connect your HR platform (Gusto, Deel, Rippling, or Justworks) in Integrations tab, or paste employee record." Write resolution, continue.
3. **Confirm authorization.** Confirm requester authorized to see confidential data for this employee. Never reveal one employee's confidential data (comp, performance, medical, immigration) to another without explicit authorization.
4. **Discover HR platform tool** - run `composio search hris` for read-only profile slug. Pull: role, level, tenure, manager, location, comp (if authorized), work-authorization / visa status (if authorized), start date.
5. **Local source pulls (read-only).**
- `onboarding-plans/{employee-slug}.md` - if this agent onboarded them. Skim Day-30/60/90 hits + misses.
- `checkins/` - scan most-recent check-ins referencing this employee-slug.
- `retention-scores/` - most recent score for this employee-slug.
- `interview-loops/{employee-slug}.md` - if past candidate, pull panel-debrief signal.
- If any sibling agent dir missing (standalone install), silently skip, note "N/A - sibling not installed" in dossier.
6. **Compose dossier** with four sections:
- **Profile** - name, role, level, tenure, manager, location, work-authorization status (if authorized).
- **History** - hire path (recruiter → offer → start), onboarding highlights, level changes, comp changes (if authorized).
- **Recent signals** - 1:1 themes from last N check-ins, retention score + trend, recent approvals run through this agent.
- **Upcoming** - next review date, visa expiry (if any), equity cliff (if any), next milestone from onboarding plan.
7. **Write** dossier atomically to `dossiers/{employee-slug}.md` (`*.tmp` → rename). Keep to one scannable page.
8. **Append to `outputs.json`** - read existing array, add `{ id, type: "dossier", title, summary, path, status: "draft", createdAt, updatedAt }`. Write atomically.
9. **Summarize to user** - one paragraph with headline signal (tenure + retention score + next milestone) and path to artifact.
## Never
- Never modify HR platform / payroll records. Reads only.
- Never invent tenure, comp, or performance data. If source missing, mark UNKNOWN.
- Never leak confidential data cross-employee without explicit authorization.
## Outputs
- `dossiers/{employee-slug}.md`.
- Appends to `outputs.json` with type `dossier`.